The present invention relates to a method and apparatus for data warehouse model validation.
A data warehouse is a central repository of integrated data, typically formed from by an extract transform load (ETL) process acting on many source systems in use within an organization. Validating the design and completeness of a data warehouse model can be challenging for the following reasons: data is not always modelled consistently in different source systems and the data warehouse may need to accommodate many data sources of varying levels of quality and consistency. Source systems may not be formally modelled (for example unstructured sources and spreadsheets) and may have minimal constraints applied to the data; and for source systems that are formally modelled, the quality of design of the source system models may differ widely, this can be for many reasons including the skills & experience of the data modeler.